Does inflation really lighten public debt?

In 2020, French public debt weighed 114.9% of gross domestic product (GDP). In 2023 it weighed only 109.5% — a fall of 5.4 points — even though the State had repaid nothing: the stock had risen from €2,663.9bn to €3,103.2bn. Then the ratio rose again, to 115.7% in 2025.

The idea that “inflation lightens debt” is widely repeated. This page checks what it means with the national-accounts figures published by the Institut national de la statistique et des études économiques (INSEE): through which channel inflation acted, in what proportion, and why the effect stopped.

A ratio has two terms. Between 2020 and 2023, the numerator — debt — rose by €439.3bn. The denominator — GDP in value — went from €2,318.3bn to €2,833.8bn, or +22.2%. Debt rose, but GDP in current euros rose faster: that is the whole mechanism.

This rise in GDP in value splits into volume and prices. Real activity grew by +11.5% over the three years (including the 2021 rebound); the rest, +9.6%, is the effect of prices on GDP — a France ÉcoScan calculation by difference. The consumer price index, for its part, gained +12.2% over the same period: the two measures of “inflation” do not coincide, because GDP is not a consumption basket.

Had GDP grown only in volume, with no price increase at all, the 2023 ratio would have been about 120.0% instead of 109.5%. The difference is the share of the ratio's decline that can be attributed, by accounting construction, to rising prices.

Annual inflation went from 1.6% in 2021 to 5.2% in 2022 and 4.9% in 2023, then fell back to 2.0% in 2024 and 0.9% in 2025. The denominator therefore stopped running: GDP growth in value is only +3.6% in 2024 and +1.9% in 2025.

Meanwhile the numerator keeps advancing at the pace of the deficits — 5.4% of GDP in 2023, 5.8% in 2024, 5.1% in 2025. A deficit of five points of GDP against nominal growth of two per cent mechanically pushes the ratio back up: that is what happened, with 6.2 points regained between 2023 and 2025.

Inflation acts first on the denominator, and immediately: GDP in value incorporates the price rise in the same year. It then acts on the numerator, but with a lag and in the other direction — indexed public spending (pensions, benefits, wages) is uprated, revenues follow prices, and the part of the debt indexed on inflation sees its cost rise. The net effect on the deficit is not measured here.

Over the long run, there is no relief: since 1990, prices have risen by +79.6% and the debt stock has been multiplied by 9.0. Inflation has reduced the real value of every euro borrowed yesterday, but it has never stopped the stock from growing far faster than prices.

The reverse episode has happened. In the early 1990s, disinflation — 3.3% in 1991, 2.1% in 1993 — coincides with the 1993 recession (−0.4%): the denominator stalls, the deficit explodes, and the ratio jumps from 37.8% to 48.2% in two years. Low inflation is no relief; combined with zero growth, it is what pushes a ratio up the most.

The 2022-2023 shock is, over the past thirty-five years, the only period in which inflation exceeded four per cent two years in a row. That is what makes the episode remarkable, and also what makes it impossible to reproduce at will: nobody chooses their inflation.

In the second quarter of 2026, the ratio reaches 119.0%, above its 2020 level. In 2025, the last closed year, inflation had fallen back to 0.9% on an annual average, far from the 5.2% and 4.9% of 2022-2023, and the deficit stood at 5.1% of GDP: neither of the two conditions for the 2021-2023 decline was met. This page serves annual inflation only; 2026 will enter it once the year is closed.